Chapter 1. Epidemiological Aspects Of Homicide | Psychiatry Online (2025)

HISTORICAL PERSPECTIVE ON RATES OF HOMICIDE

Trends in deaths from homicides for the twentieth century reveal low rates early in the century, with a peak first reached during the economic depression years of the 1930s.1 The homicide rate dropped during World War II, followed by a slight increase in the late 1940s and early 1950s. By the end of that decade, the rate had dropped, but it picked up again in the 1960s with a steady climb. The peak rate of 10.2 murders per 100,000 people was reached in 1980. During the 1980s, the homicide rate continued to be among the highest in the twentieth century in the United States. From 1990 to 1994 the rates remained quite high, ranging from 9.0 to 9.8, but then began to drop. In the last years of the century, the rates for serious felonies, including homicide, began to drop and continued to drop into the twenty-first century, with a rate of 5.5 in 2004. Such data are available from the annual publication of the Uniform Crime Reporting (UCR) Program by way of reports to the Federal Bureau of Investigation (FBI) and its Supplementary Homicide Reports (SHRs) and are summarized in Table 1–1. The debate has been why these changes occurred.

In a report by the National Academy of Sciences,2 it was noted that the annual risk of becoming a homicide victim had risen from 1 in 12,000 in 1987–1988 to 1 in 10,600 by 1990. However, the lifetime risk of homicide was much greater, with the highest lifetime rate in six demographic groups being among black males. Their rate of 4.16 is equivalent to a 1 in 24.1 chance of dying by homicide. For American Indians, the risk is 1 in 57, whereas for white males and females the risk is less than 1 in 100. It must be emphasized that despite the media’s focus on killings of adolescents and young males, less than 25% of one’s lifetime homicide risk is actually incurred by the time that person reaches his or her 25th birthday. Thus the high number of black males dying by age 24 years must be seen in the context of the high homicide rate for black males at all ages. Arrest data for 2003 indicate that males composed 89.7% of murder arrestees. By race, whites accounted for 49.1% of the murder arrestees and blacks 48.5%, even though only 12% of the population is black.

Data from the UCR indicate that a record number of 23,438 overall homicides were committed in 1990.3 However, this figure by itself is relatively meaningless; it must be viewed in relationship to other time-measured variables to have any meaning. For example, an increase in population could lead to an increase in the number of homicides. This factor has led demographers to study longitudinal trends that include homicide rates. The overall rates can vary significantly over time. In addition, the data may be broken down into rates for various states and regions of the United States and may detail the relationships between the people involved in the killing (e.g., homicide victims being spouses, children, and so forth). For example, in the early 1990s, the western United States appeared to supplant the South as the region with the highest overall level of violence among the four regions the UCR program uses. However, by 2003 the southern states, which make up 35.9% of the nation’s population, had regained the position by accounting for 43.6% of the murders in the country.4 Another important variable is the demographics, such as the proportion of those in a high-risk age group for committing homicide at a particular time.

SPECIFIC EPIDEMIOLOGICAL ASPECTS

Nine key epidemiological variables have been studied or raised as significant with respect to homicide: age, race, gender, socioeconomic class, method of killing, relationship between the victim and perpetrator (including cases that some interpret as being precipitated by the victim), the perpetrator’s prior arrests, use of alcohol or other drugs, and temporal and ecological factors.

One caveat is needed before exploring the individual epidemiological variables. Epidemiological data reflect data on the victims but not on the perpetrators. This omission is interesting in its own right and perhaps signifies that understanding the mind of the murderer requires a clinical or criminological explanation. Thus, in an operational sense, homicide rates can only give the numbers of people killed in any given population unit.

Many of the factors discussed in this section overlap, and no scientific presentation can make a claim that any one variable is the cause of a homicide. Instead, a workable model for determining factors leading up to a homicide must elicit significant background factors that under certain precipitating circumstances, predispose an individual to commit a homicide. The inductive search for determining factors must be an empirical one in which the basic premise is that for any one homicide, several factors are necessary but one factor is never a sufficient cause for that homicide. Such a position is compatible with considering the significance of personality attributes and given clinical diagnoses. It is also consistent with the position that when legal questions arise regarding a person’s responsibility for a homicide, no answer can ever be given simply on the basis of a diagnosis. Overdetermination of homicidal actions is the rule.

Race

The racial factor is reflected in fluctuations over time in general, but again, rates for nonwhite homicide victims and perpetrators are higher than for white victims and perpetrators for every age group. Depending on the age group studied, the rate for homicides among nonwhite males can be as high as 10 times that for white males. It should be noted that most studies have focused on larger American cities. In a study of homicide in New York City, Tardiff and colleagues6 found that young black and Latino men were more likely to be homicide victims than men from any other demographic group. In the subset who died within 48 hours, 31% were positive for cocaine metabolites, and 25% of their deaths involved firearms, which suggests a basis for the high death rate. In turn, these variables may be situational and related to poverty.

The significance of race in epidemiological studies of homicide cannot be ignored. Yet the deeper question is, what does it mean? Besides the national data noted earlier, other studies have documented that black people are involved at a rate far in excess of their numbers in the population. The studies are also consistent in indicating that such killings are overwhelmingly intraracial. Findings from Wolfgang’s7 early study in Philadelphia, PA, reported that based on 1948–1952 data, 73% of the offenders and 75% of the victims of 588 homicides were black; these findings were consistent with a later study by Hewitt.8 Furthermore, the distribution seems to be continuing even though the overall murder and manslaughter rates have declined from the early 1990s.

These data are depressing and unremitting, and they cry out for an explanation as well as an amelioration of the situation. Young black males and their families take the brunt of this public health problem. When it is realized that blacks account for only about 12% of the U.S. population but disproportionately account for approximately 50% of those arrested for murder and 35% arrested for assaults, obviously there is a problem that has so far remained unsolved. In terms of rates, blacks are six times more likely to be victimized and eight times more likely to commit homicides than whites. Figures 1–1 and 1–2 illustrate the different rates of victimization and offending of blacks and whites for homicide in 2002. It can also be noted that in terms of the years of potential life lost, homicide exerts a greater effect than many other causes of death, including suicide, because homicide victims are from a younger group on average. This in itself would make it a major social problem.

Several explanations have been offered for these findings. Traditional explanations have raised the possibility of social and economic inequality, lack of opportunity, racism, and discriminatory practices in the criminal justice system, ranging from the point of arrest to sentencing. Some have argued that such disparities exist for minorities in diverse cultural settings outside the United States as well. Yet although all these factors are relevant, in themselves they do not dispose of the question. To complicate matters, the high-risk group of black males—those in the 14–24 age group—has changed very little in terms of numbers from 1993 to the present overall decline in homicide rates. It is interesting to contrast these data with scholarly studies of why long-term homicide rates have declined in Europe. Explanations are offered in terms of a decline in elite homicide (i.e., homicide by the upper social classes) over a long period and a drop in male-to-male conflicts in public space, but the racial factor is omitted, although noted age and sex patterns remain relatively unchanged.9

Consider the explanation of economic and social deprivation. Both topics are raised as a correlate of homicide that can have some possible causal significance or possibly none at all. In consideration of how these factors may operate, the concepts of poverty as a state of absolute deprivation and inequality as a relative deprivation are relevant. The result is that fewer economic resources are available for those in such a predicament. If the focus is on the relationship between poverty and homicide rates in large American cities, there is a consistent and positive correlation to all homicide rates. What is important is that this correlation exists independently of variables such as race, region, or population size. However, various kinds of inequality do not show an effect on homicide rates, also independent of these same variables.10

RELATIONSHIP BETWEEN VICTIM AND PERPETRATOR

Although the victim–perpetrator relationship in homicides is often one of family ties or intimate relationships, the number of homicides committed by strangers has increased significantly in recent years. As characterized in the UCRs, the majority of homicides occur in the context of what is referred to as primary homicides, with primary referring to a death that occurs in a relatively spontaneous fight between people who know each other, often over money or property and fueled by alcohol or drugs. According to the 2004 UCR, 12.5% of perpetrators and victims were members of the same family and another 30.5% were friends, neighbors, and acquaintances. Only 12.5% of homicides occurred among strangers. However, in 44.5% of these cases, the police could not determine whether a prior relationship existed between the perpetrator and the victim. This may be because they cannot discover who the perpetrator was or because there is a lack of evidence to make any connection. There is also a gender difference in victim–offender relationships, in which female victims are more likely to be killed by an intimate or family member. In contrast, male victims are more likely than female victims to be killed by acquaintances or strangers (Table 1–4). If the stranger and undetermined relationship categories are combined, these ratios amount to 51.8% for males and 36.6% for females.

It is thus probable that more people are being killed by strangers than ever before. Data from the Bureau of Justice Statistics indicate that the number of homicides in which the circumstances are unknown has tripled since 1976.16 On the other hand, homicides involving adult or juvenile gang violence increased more than sixfold in that time period. In the past, there has been a customary 70%–75% “clearance rate” for homicides, where the police believe they have solved the crime and one or more persons should be arrested. The clearance has been associated with a high rate of reporting, because few homicides are never reported to authorities. By 2003, the clearance rate had decreased to 62.4%, presumably related to the increase in stranger killing. It appears that about 50% of homicides stem from an argument, whereas another 25% occur under the influence of alcohol or drugs.

This discussion leads to the broader topic of how an epidemiological approach conceptualizes perpetrator–victim relationships and their potential for heightening the possibility of a homicide. One approach has been to use a victim-precipitation model, which has been criticized as blaming the victim. An alternative approach has focused on the diverse and complex interactional variables in relationships or situations that can have a violent ending.

The victim-precipitation model looks at the ways homicide victims played some role in their own future deaths, focusing on situations such as arguments in bars or disagreements between spouses and lovers that escalated into increasing taunts, insults, or physical contact. When the situations were later reconstructed from the available facts, it often seemed that what had pushed matters toward homicide often began with a trivial event or exchange. The presumption in such theorizing is that the individuals, although not necessarily intimates, were interacting with each other, and therefore the killing would not be seen as part of a model for killing by a stranger. This model would thus not be applicable to killings occurring in the context of robberies or burglaries.

In an interesting approach that parallels many psychiatric theories of interpersonal violence, criminologists look at the types of interactions between individuals that eventually lead to a death. The difference between this theory and earlier theories about victim-precipitated homicide is the realization of how a homicidal outcome is a “situated” transaction that could easily have many possible outcomes. Frequently, a situation escalates into violence because neither participant allows the other the opportunity to back away while still saving face. These escalations are neither the stuff of dramatic television portrayals nor the act of a deranged killer who could later plead not guilty by reason of insanity. Instead, these types of killings arise in the give-and-take of ordinary transactions between people and comprise a frequently found framework in which homicide is enacted.

Miethe and colleagues17 conducted a study of situational contexts in homicides, focusing on the interactions among offenders, victims, and elements of the offenses. They relied on data from narrative accounts and the SHRs over three decades using a methodology of qualitative comparative analysis. One of their findings was that categorizing homicides as either instrumental (those carried out for an explicit future goal, such as acquiring money or status) or expressive (unplanned acts of anger, rage, or frustration) is excessively simple. There is so much variance among the nature and prevalence of the offender, victim, and situational factors that the distinction between instrumental and expressive motives is unclear. The configuration of both is required to begin to understand the social processes and triggering mechanisms involved when individuals and situations converge for a homicide.

Psychiatry’s major contribution to comprehending how these seemingly commonplace behaviors lead to homicides is in terms of understanding personality vulnerability and functioning and the impact of painful emotions on human beings. These painful emotions are evoked in the course of interactions in which sensitivities or misinterpretations are played out. In some situations, certain personality disorders or traits predispose the individual to overreact or misinterpret. For these people, the sway of emotions then surges to the surface and erupts into some impulsive action. The actors might be spouses or lovers who argue and then progress to shoving and slapping and then to the use of fists, objects, or weapons. Name-calling or taunts are frequent. In some cases, it might be a child’s disobedience or the endless crying of an infant that elicits feelings of helplessness in the adult, who then reacts to the child’s behavior as a provocative challenge that must be handled. At a crucial juncture, the point of no return is crossed.

Luckenbill,18 a criminologist, broke down these various types of interactions into a series of stages. In stage one, the future victim performs some type of act or verbalization that is interpreted as an attack on the future perpetrator’s activities or self-esteem. Seen from a psychiatric perspective, this initial stage would reflect the perpetrator’s tendency to distort or overinterpret. For example, an individual with paranoid or borderline personality would be among those likely to overreact. Stage two is an attempt to confirm the meaning of what the potential perpetrator has heard. Other people may be turned to for their interpretation; in fact, it is interesting that a majority of such confrontations occur in the presence of others. By the time stage three is arrived at, the perpetrator is at the point of responding, often by challenging the intended victim or demanding a halt to or retraction of certain statements. Threats may emerge. Stage four is the counterresponse of the intended victim, who at that point does not back off but stands his or her ground. Stage five is the emergence of violence, either abruptly or after the perpetrator seeks out and obtains weapons. Stage six involves the reaction of the perpetrator after the killing, such as fleeing or being restrained by others.

In this framework, many variables are operating, and indeed are necessary, for the predicted final outcome to occur; these variables involve both instrumental and expressive motivations. One necessary element is the occurrence of an intense interaction in which others are usually present but in which no one intervenes. The ready availability of weapons as a part of a high-risk lifestyle may also be an operating variable. Such a milieu is typically present in places such as bars, entertainment clubs, or street-corner confrontations. Note that through the first four stages of this model, various outcomes are possible.

PERPETRATORS’ PRIOR CRIMINAL OFFENSES

There are many difficulties in assessing the significance of prior delinquent or criminal behavior with respect to the later commission of a homicide. A major problem is related to risk assessment or the prediction of such behavior. There is the further complication of moving from the realm of early adult criminal behavior to also encompass the antecedent role of adolescent antisocial behavior and whatever its relationship is to later violence in the form of becoming a future murderer. Specifically, much difficulty is attendant on predicting rare events such as homicides. When a retrospective look is taken after a homicide has occurred, certain variables often appear prominent. However, such retrospective approaches do not allow much sensitivity when it comes to antecedent prediction. A practical problem is that juvenile files are sealed in most states unless the juvenile has been waived to the adult criminal justice system. At that point, and for most young adults, it is usually not their first encounter with the legal system.

Such data are not to be interpreted as indicating that most juvenile offenders go on to become violent adult offenders. This is not the case. The data indicate that of the 33%–45% of males in the United States who were once detained by the police or arrested as juveniles, the majority (approximately 65%) were arrested only once or twice and have no subsequent adult record. What these data mean is that a small number of juvenile offenders persist in committing crimes into young adulthood and that those juveniles who engaged in serious crime at an early age form a subgroup of individuals who are more likely to engage in serious violence, such as a homicide, in their 20s.

The highest rates for victimization and perpetration of a homicide occur among older teens and young adults. The rates for these groups increased dramatically in the late 1980s and early 1990s and then declined, as they did for all ages, as shown in Figures 1–3 and 1–4. In the period from 1985 to 1992, the homicide victimization rates for 14- to 17-year-olds increased almost 170% and then declined to levels similar to those experienced between 1976 and 1985. The highest victimization rates in that period were in the 18- to 24-year-old age group, which replaced the 25- to 34-year-old group. Offending rates of 14- to 17-year-olds increased rapidly after 1985, surpassing the rates of 25- to 34-year-olds, but the latter group has now recovered its higher position over the teenagers. The 18- to 24-year-old age group has the highest homicide rate, and from 1985 to 1992 their rates almost doubled; the rates declined since then but still remain higher than their levels prior to the mid-1980s.

At the height of these killing rates, the idea of a group of juveniles and young adults who were “superpredators” was broached. The implication was that this would be the group from which future killers would emerge. However, it was realized that many of these killings were not planned acts of homicide but rather impulsive acts often connected with drug dealing. They were thus often gang related and carried out by two or more individuals. Much of the impetus for the killing came from the combination of drug dealing and the easy availability of guns. The phenomenon of group identification and the idea of needing to join in group activities or losing a reputation was often at stake. There have been endless discussions about what contributed to the lowering of rates in this age group. Suggestions included the shift away from crack cocaine, police adopting a mixture of new tactics varying from cracking down to fostering more community involvement, and schools taking on monitoring roles such as checking for weapons. However, the rise and fall of gangs will inevitably continue, as witnessed by gang-related killings in large cities again increasing to a level double that of 1999.19

The early onset of delinquent behavior is often seen as a good predictor of later chronic, serious, and violent criminal acts when the youth becomes a teenager or young adult. When a juvenile’s early offenses include serious violent offenses (e.g., robbery), his or her probability of committing serious crime as an adult is increased. A third predictor variable in the juvenile offender is repetitive property offenses. Delinquency convictions before age 16 years and repeated sentences to juvenile correctional facilities are two other significant variables. When these variables are taken together in a weighted summative manner, the probability increases that a juvenile offender will commit a violent offense—including an increased possibility of committing a homicide—as an adult. In Wolfgang’s study⁷ of 588 homicides, two-thirds of the perpetrators had prior arrest records. It was revealing that most of these arrests had not been for trivial offenses, nor even for property offenses, but for offenses against people that usually had been violent.

Meta-analysis has been used to try to get at predictor variables in longitudinal studies of antisocial behavior as related to later violent behavior. Although statistical significance is not the same as practical significance in identifying who is at risk for violent behavior, it is known that only a small number of individuals will engage in homicide-level violence in adolescence or early adulthood. Meta-analysis allows small samples to overcome the low base-rate problem. By this approach, the strongest predictors for later serious violence beginning at ages 6–11 years were committing general offenses and engaging in substance use (chiefly tobacco and alcohol). A second-ranked group of predictor variables was male gender, family socioeconomic status, and antisocial parents. The strongest predictors at ages 12–14 years were lack of social ties, antisocial peers, and committing a general offense. A third-ranked set of variables was aggressive behavior, school attitudes and performance, various psychological conditions, parent–child relations, and physical violence. The conclusion for these groups is that these characteristics can distinguish juveniles who are at risk for violent, serious offending later in the 15- to 25-year age range.20 Of course, such data can never specify who is going to commit a homicide, but only the group who is at higher risk.

The psychiatric significance of these data is that they indicate that homicide perpetrators had unresolved conflicts involving family problems and aggression and that there had been a failure to contain that aggression. It can be hypothesized that these conflicts are replayed in later acts of violence against individuals that culminate in a subsequent homicide. Although criminological researchers have not studied just what these psychological or social conflicts might be, this hypothesis points to the need for parallel lines of inquiry. It would also be significant to determine which intervening variables contributed to the remaining one-third in Wolfgang’s study⁷ later committing a homicide or whether these individuals were just “lucky” that earlier personal violence did not lead to arrests.

ROLE OF PSYCHOACTIVE SUBSTANCES

Alcohol or other psychoactive drugs are mentioned so frequently in connection with various incidents of violence, including homicide, that the specific role they may play is often not discussed. In fact, the words alcohol and violence are often used together as though they are inseparable, with the implication that alcohol causes the violence. This belief has become almost dogma among many in top political positions and has been reflected in legislative and community decisions involving fiscal distributions. The preconception is that alcohol facilitates the release of aggression. This position is usually based on data and figures that indicate a high percentage of violent acts are perpetrated by individuals under the influence of alcohol, even though in many cases the blood alcohol level of perpetrators was not taken, and the connection was made solely based on observation. The same thinking applies to other drug usage, in which, despite the varying effects of different drugs, the suggestion is that the person imbibed or took a drug shortly before a homicide occurred. Accurate tests to confirm or disconfirm the connection between homicidal violence and alcohol use would need to compare data between perpetrators who have ingested alcohol or psychoactive substances and individuals who have ingested similar substances but have not acted in this manner; these tests would also require control of variables such as when and where the drinking or drug use took place. It is easy to bypass other significant factors involving drug use besides the psychopharmacological effects. There is the resort to crime to get money to buy drugs, which increases the chance of violent encounters. In addition, the drug dealing trade may involve competition and intimidation techniques that can lead to homicides. When such dealings involve gangs, the potential for homicides is magnified. Drug abuse, especially alcohol, is also involved in many vehicular homicides.

In the absence of any definitive studies of these factors, a large number of studies suggest a relationship between psychoactive substances and homicide on a biological or sociological level. In Holcomb and Anderson’s study21 of males charged with first-degree murder in Missouri, the authors found that 55% of the perpetrators had ingested drugs, alcohol, or both at the time of the killing. Mayfield22 found that 57% of convicted murderers studied had been drinking at the time of the killing. Greenberg23 reported that alcoholism was diagnosed in 20%–40% of convicted murderers. A similar figure was noted in a California study of females arrested on homicide charges: 51% had been drinking at the time of the killing.24 Some 60% of violent offenders are estimated to have consumed alcohol at the time of their offenses.25 Roizen26 found that 86% of homicide offenders had been drinking at the time of the offense. Results from such studies support the generalization that about half of homicides occur under the influence of drinking. What the studies do not bring out sufficiently is that in many of these situations, the perpetrator and victim have both been drinking. In Wolfgang’s study,⁷ only the perpetrators had been drinking in 11% of the homicides, only the victims had been drinking in 9%, but both the perpetrators and victims had been drinking in 44%. Subsequent studies support a figure of close to 50% for the percentage of victims who had been drinking at the time of their deaths.27

However, despite the common assumption that alcohol, in amounts that are not sedative, increases one’s potential for aggression, the action–reaction relationship is not linear. The outcome depends on a host of variables, such as the cultural setting, genetic vulnerability, neuroendocrinological factors, other drugs that may have been taken, and immediate precipitants. Hence, the mechanism for how a violent end-point occurs is not entirely clear. Some studies find little evidence of a simple direct relationship between the psychoactive effects of drugs or alcohol and aggression.28 The pattern of drinking may play a role, such as in the more concentrated style of drinking prevalent in the United States rather than the spaced drinking style prevalent in some European countries.

The suggestion of a type of alcohol-facilitated, episodic dyscontrol syndrome arises. Yet results from studies in Scotland29 and the Bordeaux region of France30 have indicated that 55% and 51%, respectively, of those convicted for murder had been drinking at the time of their offense.

One persistent source of controversy is the extent to which someone carrying out a homicide under the influence of alcohol is responsible for that act. The answer to this question, in turn, is connected in part with a belief system, neither proven nor disproven, that alcoholism is a disease. The inference is that individuals who are biologically vulnerable to the effects of alcohol should not be seen as responsible for their violent acts committed under the influence. This type of reasoning has often been subjected to critical appraisal.31

In addition, not everyone subscribes to the theory that the loss of inhibition accounted for within a biological disease model of alcohol use is the crucial variable in assessing personal responsibility for behavior. Instead, proponents of a social-cognitive model have focused on the learned expectations of what alcohol supposedly permits one to do. Thus if the expectation is that when someone drinks, his or her control over personal impulses will be lost, then it is more likely that he or she will lose control. Bartol¹¹ summarized the debate from this social-cognitive perspective, in which the consumer’s expectations of the effects of alcohol act as a crucial variable in how he or she will behave when drinking. Those holding to the social-cognitive model see the psychoactive influences of alcohol as learned and situationally determined and thus see the loss of control as related to a cognitive expectancy rather than a disease.

On the other hand, no one can doubt that alcohol has physiological effects on the nervous system. A correlation exists between low cerebrospinal fluid concentrations of 5-hydroxyindoleacetic acid (5-HIAA), a serotonin metabolite, and impulse control problems.32 Impulsivity and disinhibition have consistently been related to individuals with chronic alcoholism as well as related to Cluster B personality disorders.33 Also, the results of genetic work on alcoholism and antisocial personality have raised questions about whether there is some overlap among those who carry one or both diagnoses.34 There also appear to be differences between males and females regarding the connection between alcohol consumption and violent behavior that may be related to different social expectations.35 In this connection, there is an observation that men using cocaine are more likely to perpetrate violent crimes, whereas women using cocaine are more likely to become victims of violence.36

The debate concerning alcohol use and violence may primarily be over what is the prime influence on the violent behavior and whether alcohol is the necessary variable in an event culminating in violence. In a carefully designed research project involving all violent acts witnessed by or perpetrated on a random sample of citizens in a single town, Pernanen37 found that alcohol was involved in 54% of those acts but that the effect of other factors on the violent outcome varied even more than alcohol use by the victims and perpetrators. Some studies have demonstrated the missing effect by giving alcohol or substances purported to be alcohol to study subjects under single-blind conditions.28,38 The researchers then observed the effect of the subjects’ believing they had drunk alcohol as well as the context in which that alcohol had been drunk on the subjects’ behavior. There are limits on how far the cultural setting can influence behavior if alcohol continues to be consumed. Eventually, higher cognitive processes (e.g., attention, concentration) do become impaired. In the end, it is the combined effects of the pharmacological, situational, and cognitive factors that determine the outcome: feelings of being threatened, provoked, or insecure; intrapsychic components (e.g., depression, anxiety); and the influence of alcohol can all contribute to an act culminating in homicide.

A similar critical appraisal could be done for the effects of psychoactive drugs on an individual’s potential for extreme violence. In one study, 81% of 275 homicide victims were tested for benzoylecgonine, the major metabolite of cocaine. Forty percent of the victims tested positive for this metabolite, a number much higher than that presumed in earlier studies.39 Although marijuana or opiates may inhibit violent behavior, phencyclidine, lysergic acid diethylamide (LSD), and amphetamines are seen as increasing the potential for violence. The changing patterns of drug usage and drug subcultures, especially in major American cities, need to be assessed at any given time. Drugs vary in popularity at different periods of time. For example, in the 1970s heroin injections were popular, followed by the emergence of cocaine. In the 1980s and 1990s, smoking crack cocaine emerged with its accompaniment of increased violence. In the 1990s, the increased use of methamphetamine began, and it continues into the twenty-first century. A related drug that has emerged is methylenedioxymethamphetamine (MDMA), which is known as “Ecstasy” and is a stimulant and psychedelic that induces emotional lability and surges of energy so a person can stay active for prolonged periods of time. Diverse drugs are still used and different patterns witnessed, such as smoking marijuana in a cigar (a blunt) along with consuming alcohol. All of these have different potentials for violence depending on the vulnerabilities of the individual, the social context, and whether the person is also involved in dealing drugs.

TEMPORAL AND ECOLOGICAL FACTORS

CULTURAL VARIABLES

In addition to discussions of different homicide rates within different social groups in a particular society, it must be recognized that cultural settings also influence homicide rates. This is true even when making comparisons among industrialized countries, let alone between industrialized and less industrialized countries.

Within the industrialized nations, the United States has the highest murder rate; Japan and Austria have the lowest rates. Some economically developing nations do have higher homicide rates than the United States; the homicide rates per 100,000 in these countries are as follows: Colombia (56.3), El Salvador (38.8), Jamaica (37.2), and Mongolia (24.6); however, it is not clear whether these rates exclude war-related activities.43

The homicide victim rate for most industrialized countries is usually below 3 per 100,000. However, from 1980 onward, the homicide rate in the United States has been more than twice that figure, falling somewhere between 5 and 10 per 100,000. When the rates are broken down in terms of age groupings, the disparity in the rate in the United States compared with most European countries becomes even more glaring, because the rate for young males in Europe lies between 1 and 2 per 100,000. As far as can be determined from existing documentation, the murder rate in the United States in the twentieth century has never been as low as 3 per 100,000. It is informative to compare the numbers of homicides in different countries. Interpol data from 1999 to 2000 show rates of homicide per 100,000 for Europe to be 1.59.44 Certain countries stand out, such as Lithuania (10.62), Estonia (10.61), and Latvia (6.47). Russia is listed at a rate of 22.05 and South Africa at 55.86. At the lower end are Spain and Switzerland (1.12), Sweden (1.11), Japan (1.05), and Norway (0.05).

Explanations of this striking disparity vary widely. One is that Americans are more tolerant of aggression and violence, including violence with firearms, which are considered to be ever present in American society. (See earlier discussion of gun-related homicides in the section “Method of Killing.”) Another theory is that American society does little to discourage physical aggression among its young male population. Still other theories stress the effects of the media and network television violence on America’s tendency to be a violent nation. The conclusion of some researchers is that the United States is a society that promotes and accepts the extremes of personal aggressive violence but is hypocritical about other standards, such as sexual deviance, as witnessed in discriminatory treatment of those with diverse sexual practices.45

SOCIOLOGICAL VARIABLES

SUBCULTURE OF VIOLENCE

The essence of the theory of a subculture of violence is that certain subgroups in American society hold values conducive to violence and that the socialization process in those groups not only accepts but promotes such behavior. The initial formulation of this theory was not restricted to homicide but rather was applied overall to criminal behavior in which such behaviors were seen as being learned. Through interactions with individuals who approve of crime, norms develop that are conducive to crime within certain subgroups. For example, a subgroup may be more tolerant of violence and likely see it as a legitimate means for the expression of aggression or violent tendencies. If a homicide results, it is seen as unfortunate for both parties but simply an inherent risk in the way life is lived in the subculture. Such an outcome is not seen simply as two individuals interacting but as a reflection of some social groups more often resorting to violence that in turn increases the likelihood of homicide.

The exposure of members of these subgroups to violence can be intense, from the enormous violence witnessed on television and in films to the subculture’s connection of male identity with expressions of violence. Weapons are readily available for those who believe they need to “even things up” to express their aggression. Going along with the acceptance and promotion of these values is the hypocritical policing of sex in magazines, music, and television in society, which gives a strange mixture of messages in which sex is seemingly regulated but violence promoted.

One hope for members of this subculture is that the violence will be channeled into other options (e.g., competitive athletics among these groups). Another hope is that those who are observers and vicarious participants in the subgroups can mitigate the degree of their own need for direct aggression.

Of course, the linchpin of the theory that a subculture of violence leads to increased homicides is that those subgroups who have high rates of homicide actually do possess a different set of norms with respect to violence. The reverse—that high homicide rates validate such a culture—cannot be argued; rather, independent empirical validation of the theory is needed. Measuring attitudes toward violence is difficult; even then, expressed behavior may not match verbalized values. For example, almost no one directly states that he or she approves of violence, let alone the killing of someone (with the exception of those involved in political killing and ethnic cleansing). Rather, it is the covert way in which different attitudes toward violence are expressed that would distinguish one group from another. As a result, it is not easy for groups or individuals to be classified as part of a subcultural group or not. It is more likely that they could be classified along a continuum. One group may approve of a good deal of criminal behavior and stand in opposition to many societal values, yet only approve of certain selected killings. In between may be groups that have a mixture of values, some of which may be conducive to violence and others not. On the other end of the continuum may be a group that has values approving of more minor antisocial behavior, such as white-collar or bureaucratic violations, but that sees a homicide that results from such activities as outside the accepted norms. Such continua exist for groups as well as individuals within the group.

FELONY MURDERS

In some epidemiological studies, the concept of primary homicide (again, meaning a death resulting from a more or less spontaneous confrontation or fight between individuals or even groups) has been distinguished from the category of felony homicides, meaning killings taking place in the context of another felony (e.g., robberies, sexual killings, executions, violence between rival gangs, murders planned to achieve a certain goal [one gang sending a “message” to a rival gang]). The English common law was originally described as holding felons strictly liable for all deaths caused in the course of felonies. This rule became the rule by default in American jurisdictions and has been described as a myth.57 However, various state statutes were enacted in the nineteenth century that began to incorporate these ideas, although with different principles of liability for a felony murder. Although the number of felony homicides is significantly less than the number of primary homicides, with some estimating their prevalence at only about 20%, they continue to receive much more attention in the media than primary homicides because of their dramatic aspects and because they have greater appeal in terms of being fictionalized. When homicide rates began to increase in the 1960s, the proportion of felony homicides also increased. The result was a drop in the clearance rate for homicide because of the increase in these types of felony homicides, which do not have as high an arrest rate.

Those who have studied felony homicide often see it as an impersonal and predatory type of murder in contrast to murders committed in the heat of passion by intimates. In a classic work discussing felony homicides, Dietz58 listed their distinguishing characteristics:

Robbery homicides are the most frequent type of felony homicide; a bank robbery is only one example of this. Dietz hypothesized that robberies occur in distinct stages. First is the planning stage, which usually occurs within a few days of the actual robbery. Sometimes those planning the robbery may specify killings during the robbery, but in other cases not all those to be involved in perpetrating the robbery are informed of this aspect of the plan, whether it is conceived that a killing will occur as a direct consequence of the plan or only as the need arises. One pattern may be that two or more of the group members are informed about the intended killing or make the decision they will do so if needed, with the other members acquiescing in such a decision. The victim or victims are usually regarded as someone impersonal or someone undeserving of respect. During the posthomicidal period, the members may talk about a killing that occurred in the course of a robbery and, in some cases, joke about it.

Other variations on felony homicides occur within the context of planned executions, contract killings, or revenge killings and are related to street gangs or planned killings between rivals in organized crime. In some of these situations, a professional, referred to as a “hit man,” is hired to do the killing.

Although technically considered felony murders, homicides committed during a sexual assault (see Chapter 10, “Sexual Homicide”) differ in many aspects from these types of felony homicides.

It should be noted that the concept of felony homicide has also acquired a series of legal meanings connected with cases of who might be tried for a homicide. As this concept was initially conceived, the accomplice of the perpetrator of a homicide that occurred during a bank robbery would also be charged with homicide. That concept has now been expanded so that the person driving the getaway car from a robbery in which a homicide was committed, even though he or she was not present at the killing, would be legally charged with felony homicide. As a further expansion of this concept, if a bank security guard kills one of the bank robbers in the course of the felony or attempted felony, the other members of the group attempting to carry out the felony could be legally charged with being responsible for that death.

Apart from sexual killings, psychiatric issues are much less frequently raised with this group of felony murderers than in the primary type of homicide among intimates. Those involved in organized gangs—whether juvenile gangs or more advanced criminal organizations—or those controlling the gang do not wish to explore the intricacies of their psychiatric state at the time of the offense. Such inquiry is not part of their group ethos. Second, their own self-concept is such that they do not wish to be thought of as being deviant in a psychiatric sense, which would be unacceptable, even though social deviance is acceptable. Hence, many of the complex dynamics involved in their lifestyles are never explored.

MULTIPLE-VICTIM MURDERS

Mass and serial murders have gained increasing publicity in the United States. Whether they have in fact been more prevalent in recent years, or whether American society is now simply becoming more aware of these types of killings through improved detection and investigation, is not clear. The Bureau of Justice classifies three types of multiple-victim murderers: serial murderers, mass murderers, and spree murderers.

WHY HAVE HOMICIDE RATES FALLEN?

Historically up until 1965, the homicide rate in the United States was less than 5 per 100,000 population. Over the next 20 years, it increased to a range of 8–10 per 100,000. The homicide rate then declined for a few years, but from 1985 it climbed precipitously, led by juveniles and young adults. This increase was often attributed to the crack epidemic and violence associated with drug distribution. Some of the demographic variables believed to be associated with the increase have been discussed. In the latter part of the 1990s, the rates began to drop to levels between 5.0 and 6.0 in all regions of the United States, with some exceptions in the West. The debate is why this drop occurred. Note that this homicide rate is still far above the rates in most European countries, where the rate has fluctuated between 1.0 and 2.0.⁴⁴

Theories to explain the drop abound, and they illustrate the diverse influences that have operated.76 Frequently mentioned is the aging population of baby boomers that reached middle age and were beyond the most violence-prone years. The changing scene in many major American cities has also been noted, such as the decrease in barroom killings that were more prevalent in factory towns, especially on payday. The culture of bars itself changed, with more affluent people and females being present. Bar owners always had the power to regulate patrons who drank to excess, but law enforcement began to tighten. Police departments in large cities claimed a major role in lowering the homicide rate because of their more aggressive and imaginative policing. Their approaches varied from the redeployment of police officers to confronting groups of young males loitering about street corners. This approach has sometimes involved community policing, in which a closer relationship between police officers and their precinct populations is fostered. This viewpoint has been criticized for ignoring the effects of increased rates of incarceration.77

Tripling the nation’s prison and jail populations to over 2 million by 2005 is another argument offered on why homicide rates were lowered. The idea is that if people are taken off the streets who might subsequently have committed a murder because of their criminal propensities, the homicide rate will be lowered. By 2004, 10% of all inmates in state and federal prisons were serving life sentences, which cost states about $1 million per inmate serving out the term.78 New gun-control statutes in some jurisdictions may have played some role in reducing homicides, although a countertendency in some states was to pass concealed-gun legislation that could work in the opposite direction.

The waning of the crack cocaine epidemic also has been seen as lowering the homicide rate because of the lessened effect on individual users and the reduced drug market, in which young dealers were often armed with semiautomatic weapons. Selling the drug may have seemed like an easy way to make a lot of money, but the ravages of crack became more visible in time. However, in some areas of the country, methamphetamine use has increased among the younger population, along with Ecstasy, to take its place.

A factor that has been ignored until recently as contributing to the decline in homicide rates is the improvement in medical care. This may be an offspring of bringing techniques from the battlefield to American cities. Trauma centers and the use of 911 telephone lines or cellular phones are combined with helicopters to evacuate the wounded to a medical facility, much like under combat conditions. A deceiving element lurks in this phenomenon: the number of shootings per se may not be decreasing, but rather, more people are surviving after a shooting. In fact, the number of cases of aggravated assault has not decreased dramatically but remained rather stable.

An ingenious and controversial economic theory to explain the decline in homicide rates that occurred by the end of the 1990s has been offered on the basis of the legalization of abortion.79 The explanation offered is that in 1973 the U.S. Supreme Court legalized abortion in Roe v. Wade, whereas abortion had only been available in five states until then. The result was a dramatic increase in abortion numbers from 750,000 to 1.6 million by 1980. The idea is that very often these women were unmarried, adolescents, or poor, in which case the children would have been in a group with the highest potential rates for criminal activity. Thus, in the 1990s the groups that were aborted would have been in their late teens when the rates began to drop. It was noted that the crime rate started falling earlier in the five states that already had legalized abortion and that thereafter the states with the highest abortion rates saw the biggest drop in crime, with other factors being controlled.

Lowering of domestic killings needs to be noted as another factor in lowering homicide rates. Society became less tolerant of domestic violence, and many states required police to make an arrest if there was any evidence of physical harm to either party. The Bureau of Justice Statistics data show that since 1976 the number of men murdered by intimates (spouses, ex-spouses, boyfriends and girlfriends) dropped by 68%. The number of women killed by intimates was stalled for two decades but then declined after 1993–1995 and has remained stable into the twenty-first century.

It would be unduly optimistic to think that all these variables will remain constant and that no new factors will arise related to homicidal behavior. Any shift in the variables mentioned can upset the recent decline as well. The cyclic nature of homicides over time is what has seemed most consistent in the long run.

REFERENCES